On the Complexity of the Bipartite Polarization Problem: from Neutral to Highly Polarized Discussions